How to write a dissertation
What each chapter is for, how long it should be, the order to write them in, and the structural mistakes that cost the most marks.
A dissertation is not a long essay. An essay makes one argument; a dissertation reports a piece of research and then argues about what it means. That difference drives the structure, and students who treat it as an extended essay usually produce something that reads well and scores poorly.
The standard UK structure
| Chapter | Its job | Share of 12,000 words |
|---|---|---|
| Introduction | Establish the problem and state the question | ~1,200 |
| Literature review | Show what is known and find the gap | ~3,000 |
| Methodology | Justify how you investigated it | ~1,800 |
| Results | Report what you found, without interpreting | ~2,400 |
| Discussion | Interpret it against the literature | ~3,000 |
| Conclusion | Answer the question and state the limits | ~600 |
Humanities dissertations often replace the middle four with thematic chapters, and some sciences merge results and discussion. Check your handbook. The percentages still work as a warning system: a literature review at 45% of the total is a common and expensive mistake.
What each chapter has to do
Introduction
Four things, in roughly this order:
- Establish why the topic matters, briefly and with evidence rather than assertion
- Narrow to the specific problem
- State your research question or aims explicitly, so the marker can find them
- Outline the structure of what follows
The most common failure is a first page of general background about the field. Markers read dozens of these. Get to the specific problem inside two paragraphs. If the question itself is still vague, our guides to choosing a dissertation topic and writing a research proposal will sharpen it first.
Literature review
This chapter is not a summary of what you read. It is an argument that a gap exists and that your question addresses it. If it can be restructured into a list of studies without losing anything, it is not yet a literature review.
Summarising, which loses marks
Ahmed (2024) found high attrition in adolescent trials. Brown (2023) also examined attrition and found similar rates. Clark (2022) studied retention strategies.
Synthesising
Attrition in adolescent trials is consistently high (Ahmed, 2024; Brown, 2023), yet the retention strategies proposed in response have been tested almost exclusively in clinical rather than community settings (Clark, 2022), leaving their effectiveness in schools untested.
The second version does the same reading and arrives somewhere. Ourliterature review guide covers this in detail.
Methodology
Justification, not description. Every methodological choice needs a reason and, ideally, a rejected alternative.
Describing
Semi-structured interviews were conducted with 12 participants.
Justifying
Semi-structured interviews were chosen over a survey because the study sought to understand how participants reasoned about adherence rather than to measure its frequency. Twelve participants were recruited, which is consistent with the point at which thematic saturation is typically reached in homogeneous samples (Guest et al., 2006).
Include ethical approval, sampling, data collection, analysis approach and limitations. A reader should be able to repeat your study from this chapter alone.
Results
Report, do not interpret. This is the hardest discipline in the whole document, because the interpretation is the interesting part and it belongs in the next chapter.
- Present findings in a logical order, usually mapped to your research questions
- Use tables and figures for anything a sentence would labour
- Refer to every table and figure in the text; unreferenced ones read as decoration
- Report non-significant and unexpected findings too, because omitting them is a research integrity issue
Discussion
Usually the chapter that separates a 2:1 from a first. Four moves:
- What do the findings mean in relation to your question?
- How do they sit against the literature you reviewed, including where they disagree?
- What are the limitations, stated honestly rather than defensively?
- What follows for practice, policy or future research?
Point two is where marks are won. A discussion that only confirms previous work is weaker than one that identifies a contradiction and reasons about why it arose.
Conclusion
Answer the question you asked in the introduction, in plain terms. No new material, no new citations. If your conclusion introduces an idea, it belonged in the discussion.
The order to write in
Almost nobody writes these in the order they appear, and you should not either.
| Write | Because |
|---|---|
| 1. Methodology | Most concrete, and you did it recently |
| 2. Literature review | Long, and improves with time to revise |
| 3. Results | Mechanical once the analysis is done |
| 4. Discussion | Needs the results settled first |
| 5. Conclusion | Follows directly from the discussion |
| 6. Introduction | You now know what you are introducing |
Where dissertations lose marks structurally
- An oversized literature review. Usually because reading felt more productive than writing.
- Interpretation in the results chapter. The discussion then has nothing left to do.
- Method described but not justified. The marker cannot tell whether you chose or defaulted.
- A research question that quietly changes. The one answered in the conclusion is not the one asked in the introduction.
- No signposting. Six chapters that never acknowledge each other.
Two sentences that fix most coherence problems
At the end of every chapter, write:
- What this chapter established
- What the next one therefore needs to address
This review has shown that attrition is consistently high in adolescent trials but that retention strategies remain untested outside clinical settings. The following chapter sets out a methodology designed to test one such strategy in a school context.
Twelve of those sentences across a dissertation, and the whole thing reads as one argument rather than six assignments stapled together. It is the cheapest structural improvement available.
Last updated: 30 July 2026 · Written by the AssignWise editorial team ·Spotted something wrong?
Frequently asked questions
The standard UK structure is introduction, literature review, methodology, results, discussion and conclusion, with references and appendices after. Some disciplines merge results and discussion, and humanities dissertations often use thematic chapters instead. Follow your department handbook over any general template.
On a 12,000-word dissertation a common split is 10% introduction, 25% literature review, 15% methodology, 20% results, 25% discussion and 5% conclusion. Treat it as a sanity check rather than a rule: a chapter at double its share usually signals a structural problem.
Rarely the introduction. Most students start with methodology, because it is the most concrete, then literature review, then results. Write the introduction last, when you finally know what you are introducing.
Your university library usually holds past dissertations from your own department, which is the most useful source because it shows what your markers rewarded. EThOS holds UK doctoral theses. Read for structure and argument rather than copying phrasing.
Signposting between chapters. End each one by stating what it established and what the next needs to address. Most dissertations that read as disconnected are missing those two sentences per chapter rather than missing content.
Want a second pair of eyes on it?
Send us your document for a fixed, no-obligation quote. We reply within one working hour.
Get Your Free Quote